漢籍の電子的な翻刻について : 『說文解字注』の Unicode 轉寫を事例として

Abstract

The mathematical theory of institution could present general foundations for ontology-based semantic integration. The author already presented a formalization of describing knowledge on Chinese characters by using institutinalized representational systems. This formalization can model systems constructed by characters and relations between them, and present an inference system to deduce facts on characters. The aim of this paper is to present a case study of creating digital versions of historical Chinese texts. Our target text for digitization is Duan Yucai's Shuowen Jiezi Zhu, which is a Qing print of an annotated edition of Xu Shen's Shuowen Jiezi dictionary. This dictionary was compiled during the Eastern Han Dynasty and presented interpretation of the simple pictographs and analysis of complex characters. Our digital transcription uses as a character encoding the Universal Character Set (Unicode) which includes a rich set of CJK ideographs. In the transcription with the Unicode characters, we use the three standard dictionaries selected by the Unicode Consortium. Identifing characters in origital texts requires an interpretation of them. In particular, we take notice of Duan Yucai's phonological annotations of characters. Phonological information of characters is useful for presenting digital transcription of anciant Chinese texts. Our case study of digitizing Shuowen Jiezi Zhu shows the usefulness of our formalization, which can present inference on phonological information of characters from Duan Yucai's annotations.